Romeo Valentin

Romeo Valentin

PhD student at the Stanford Intelligent Systems Lab (SISL)

Stanford University

Hi, I’m Romeo

Currently, I am working on certification protocols for machine learning systems in safety-critical domains; first in aerospace, but with future plans for medical, finance, and public policy. To progress with this task, I have been fortunate to join the Stanford Intelligent Systems Lab. If you’re around Stanford and are interested in talking to me, let’s have a coffee!

My background is in computational mathematics, both theory and implementation, and mathematical optimization, which I applied to process systems optimization during my time at CMU. I have also worked in robotics, including embedded control, motion planning and autonomous navigation.

Prior to moving to Palo Alto I have lived in Aachen, Pittsburgh, Stuttgart, Zürich, and in my hometown Holtum. I love traveling and have traversed Vietnam and Peru by motorcycle (highly recommended), and backpacked through parts of Mexico (come for the party, stay for the nature and food). I am also an avid ballroom dancer and current vice-president of Stanford’s competitive ballroom team, currently dabbling in Brazilian Jiu-Jitsu, and was fencing state-champion in Westphalia (NRW) during my teens.

Check out my CV.

Interests
  • ML certification for safety-critical systems
  • Probabilistic ML, Uncertainty quantification, Causality & Disentanglement
  • Decision making under uncertainty
  • Efficient algorithms for combinatorial optimization
  • Contributions to open science and open source software
  • Meeting Donald Knuth
Education
  • PhD at Stanford Intelligent Systems Lab, ongoing

    Stanford University, USA

  • MSc in Computational Science and Engineering, 2019 - 2022

    ETH Zürich, Switzerland

  • Research visit in Combinatorial Optimization, 2018 - 2019

    Carnegie Mellon University, USA

  • BSc in Computational Engineering Science, 2015 - 2018

    RWTH Aachen, Germany

Publications

Quickly discover relevant content by filtering publications.
(2022). Instance-Wise Algorithm Configuration with Graph Neural Networks. arXiv.

Cite DOI arXiv

(2021). Pyomo.GDP: An Ecosystem for Logic Based Modeling and Optimization Development.

Cite DOI

Contact